Sr Linguist, Speech Technology and Research Lab
Sarah Bakst, Ph.D., is a speech researcher in the Speech Technology and Research (STAR) Lab. She has a long-standing interest in individual differences in speech production and speech-motor control. Recent projects at SRI have involved text-to-speech, speaker identification, and demographic bias in AI.
Prior to SRI, Dr. Bakst was a postdoctoral fellow at the Waisman Center at the University of Wisconsin–Madison, where her neuroimaging (MEG) and behavioral research investigated the role of the auditory system in regulating acoustic precision in spoken language in two atypical adult populations: second-language speakers and cochlear implant users. Dr. Bakst completed her Ph.D. in linguistics at UC Berkeley, where she studied how individual differences, such as native language or shape of the hard palate, account for fine-grained differences in articulatory, acoustic, and perceptual variability. She has a particular interest in retroflex consonants and the languages that contain them, especially Hindi and Tamil.
Dr. Bakst has an A.B. in linguistics with a secondary field in Classics from Harvard University and an M.Phil. in linguistics from the University of Cambridge.
Highlighted publications
- Bakst, S., Cobo-Kroenke, C., Lawson, A., McLaren, M. and Stauffer, A. (2022)., “Time-varying score reliability prediction in speaker identification.” In Proc. The Speaker and Language Recognition Workshop (Odyssey 2022), pp. 207-212.
- *Bakst, S. (2021). Palate shape influence depends on the segment: Articulatory and acoustic variability in American English /ɹ/ and /s/. The Journal of the Acoustical Society of America, 149(2), 960-971.
- *Bakst S, Niziolek CA. (2021). Effects of syllable stress in adaptation to altered auditory feedback in vowels. The Journal of the Acoustical Society of America. 149: 708. PMID 33514177 DOI: 10.1121/10.0003052
*Publications prior to joining SRI International
